[2603.20009] A Super Fast K-means for Indexing Vector Embeddings

Computer Science > Machine Learning arXiv:2603.20009 (cs) [Submitted on 20 Mar 2026] Title:A Super Fast K-means for Indexing Vector Embeddings Authors:Leonardo Kuffo, Sven Hepkema, Peter Boncz View a PDF of the paper titled A Super Fast K-means for Indexing Vector Embeddings, by Leonardo Kuffo and 2 other authors View PDF HTML (experimental) Abstract:We present SuperKMeans: a k-means variant designed for clustering collections of high-dimensional vector embeddings. SuperKMeans' clustering is up to 7x faster than FAISS and Scikit-Learn on modern CPUs and up to 4x faster than cuVS on GPUs (Figure 1), while maintaining the quality of the resulting centroids for vector similarity search tasks. SuperKMeans acceleration comes from reducing data-access and compute overhead by reliably and efficiently pruning dimensions that are not needed to assign a vector to a centroid. Furthermore, we present Early Termination by Recall, a novel mechanism that early-terminates k-means when the quality of the centroids for retrieval tasks stops improving across iterations. In practice, this further reduces runtimes without compromising retrieval quality. We open-source our implementation at this https URL Subjects: Machine Learning (cs.LG); Databases (cs.DB); Information Retrieval (cs.IR) Cite as: arXiv:2603.20009 [cs.LG]   (or arXiv:2603.20009v1 [cs.LG] for this version)   https://doi.org/10.48550/arXiv.2603.20009 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submiss...